In a medical X-ray diagnostic apparatus which is a typical example of radiographic apparatus, a flat panel X-ray detector (hereinafter called “FPD” as appropriate) has recently been used as an X-ray detecting device for detecting X-ray penetration images of a patient resulting from X-ray emission from an X-ray tube. The FPD includes numerous semiconductor or other X-ray detecting elements arranged longitudinally and transversely on an X-ray detecting surface.
That is, the X-ray diagnostic apparatus is constructed to obtain, based on X-ray detection signals for one X-ray image taken at sampling time intervals from the FPD as a patient is irradiated with X rays from the X-ray tube, an X-ray image corresponding to an X-ray penetration image of the patient for every period between sampling intervals. The use of the FPD is advantageous in terms of apparatus construction and image processing since the FPD is lighter and less prone to complicated detecting distortions than the image intensifier used heretofore.
However, the FPD has a drawback of causing time lags whose adverse influence appears in X-ray images. Specifically, when X-ray detection signals are taken from the FPD at short sampling time intervals, the remainder of a signal not picked up adds to a next X-ray detection signal as a lag-behind part. Thus, where X-ray detection signals for one image are taken from the FPD at 30 sampling intervals per second to create X-ray images for dynamic display, the lag-behind part appears as an after-image on a preceding screen to produce a double image. This results in an inconvenience such as blurring of dynamic images.
U.S. Pat. No. 5,249,123 discloses a proposal to solve the problem of the time lag caused by the FPD in acquiring computer tomographic images (CT images). This proposed technique employs a computation for eliminating a lag-behind part from each of radiation detection signals taken from an FPD at sampling time intervals Δt.
That is, in the above U.S. patent, a lag-behind part included in each of the radiation detection signals taken at the sampling time intervals is assumed due to an impulse response formed of a plurality of exponential functions, and the following equation is used to derive radiation detection signal xk with a lag-behind part removed from radiation detection signal yk:xk=[yk−Σn=1N{αn·[1−exp(Tn)]·exp(Tn)·Snk}]/Σn=1Nβn in which Tn=−Δt/τn, Snk=xk−1+exp(Tn)·Sn(k−1),and βn=αn·[1−exp(Tn)],where                Δt: sampling intervals;        k: subscript representing a k-th point of time in a sampling time series;        N: the number of exponential functions with different time constants forming the impulse response;        n: subscript representing one of the exponential functions forming the impulse response;        αn: intensity of exponential function n; and        τn attenuation time constant of exponential function n.        
Inventors herein have tried the computation technique proposed in the above U.S. patent. However, the only result obtained is that the above technique cannot avoid artifacts due to the time lag and satisfactory X-ray images cannot be obtained. It has been confirmed that the time lag due to the FPD is not eliminated (Patent Document 1).
Then, Inventors have previously proposed a technique in Unexamined Patent Publication No. 2004-242741. In dealing with the time lag of the FPD, this technique removes a lag-behind part due to an impulse response based on the following recursive equations a-c:Xk=Yk−Σn=1N[αn·[1−exp(Tn)]·exp(Tn)·Snk]  aTn=−Δt/τn  bSnk=Xk−1+exp(Tn)·Sn(k−1)  cwhere
Δt: the sampling time interval;
k: a subscript representing a k-th point of time in a sampling time series;
Yk: an X-ray detection signal taken at the k-th sampling time;
Xk: a corrected X-ray detection signal with a lag-behind part removed from the signal Yk;
Xk−1: a signal Xk taken at a preceding point of time;
Sn(k−1): an Snk at a preceding point of time;
exp: an exponential function;
N: the number of exponential functions with different time constants forming the impulse response;
n: a subscript representing one of the exponential functions forming the impulse response;
αn: an intensity of exponential function n; and
τn: an attenuation time constant of exponential function n;
Sn0=0; and
X0=0.
In this recursive computation, coefficients of the impulse response of the FPD, N, αn and τn, are determined in advance. With the coefficients fixed, X-ray detection signal Yk is applied to equations a-c, thereby obtaining a lag-free X-ray detection signal Xk (Patent Document 2). The above correction for removing the lag-behind part is also called “lag correction”.
Besides the above technique of Patent Document 2, there is a technique of using backlight to reduce long time constant components of lag-behind parts (see Patent Document 3, for example).
[Patent Document 1]
U.S. Pat. No. 5,249,123 (mathematical expressions in the specification and the drawings)
[Patent Document 2]
Unexamined Patent Publication No. 2004-242741 (mathematical expressions in the specification and the drawings)
[Patent Document 3]
Unexamined Patent Publication H9-9153 (pages 3-8, FIG. 1)